From mboxrd@z Thu Jan 1 00:00:00 1970 From: Roman Mamedov Subject: Re: raid1 read balance with write mostly Date: Sun, 3 Feb 2013 04:07:58 +0600 Message-ID: <20130203040758.5e2e7f3c@natsu> References: <36edd03089332985af9a97eeaf719cd3@mail1.volny.cz> <510D8A8C.5090505@hardwarefreak.com> Mime-Version: 1.0 Content-Type: multipart/signed; micalg=PGP-SHA1; boundary="Sig_/Kzv4wA3ji5o9eF+nGCQ2zwD"; protocol="application/pgp-signature" Return-path: In-Reply-To: <510D8A8C.5090505@hardwarefreak.com> Sender: linux-raid-owner@vger.kernel.org To: stan@hardwarefreak.com Cc: tomas.hodek@volny.cz, linux-raid@vger.kernel.org List-Id: linux-raid.ids --Sig_/Kzv4wA3ji5o9eF+nGCQ2zwD Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: quoted-printable On Sat, 02 Feb 2013 15:52:12 -0600 Stan Hoeppner wrote: > So under heavy read load, in this case we'd probably want the slow 3TB dr= ive > to contribute to the workload. No, it should behave as documented. -W, --write-mostly subsequent devices listed in a --build, --create, or --add c= om=E2=80=90 mand will be flagged as 'write-mostly'. This is valid for RA= ID1 only and means that the 'md' driver will avoid reading f= rom these devices if at all possible.=20 if at all possible =3D if there is any other mirror still alive, do not rea= d from this one. BTW I also run an SSD+HDD array, and after a brief testing today I did noti= ce some reads from the HDD on reading from the array showing up in iostat. So I can confirm the bug. --=20 With respect, Roman --Sig_/Kzv4wA3ji5o9eF+nGCQ2zwD Content-Type: application/pgp-signature; name=signature.asc Content-Disposition: attachment; filename=signature.asc -----BEGIN PGP SIGNATURE----- Version: GnuPG v1.4.12 (GNU/Linux) iEYEARECAAYFAlENjj4ACgkQTLKSvz+PZwgrsACePNLAyBW/J8V0mQZYMm22nElH nYIAnA3ql32kN0R3DQrHyctFQdyzN+OE =lBgX -----END PGP SIGNATURE----- --Sig_/Kzv4wA3ji5o9eF+nGCQ2zwD--